Computer Aided Systematic Reviews
Systematic literature reviews are a major form of secondary research, that is, research which uses existing research output as its basis rather that d . They are widely used in medicine and increasingly used in other areas including computer science (esp. software engineering). Systematic reviews are critical to understanding the current state of evidence with respect to critical research question (for example, are existing anti-stroke therapies effective or does test-driven development improve software quality).
Systematic reviews, even of comparatively simple sorts, are extremely labour intensive and require both significant domain knowledge and skills as well as review specific skills (such as information retrieval). For more sophisticated reviews, such as meta-analysis (wherein data from distinct studies is pooled to produce stronger conclusions), both statistical and domain expertise of very particular sorts are required.
Knowledge representation techniques could significantly improve the process in a variety of ways including helping resolve ambiguous terms, simplifying the retrieval process, and supporting the development of conceptual models of the evidence. The aim of the project is to improve the process of developing and exploring systematic reviews esp. the
through the use of knowledge representation technology such as ontologies (esp. those in the Web Ontology Language or extensions) or Linked Data.
We currently have a pilot project with the National Institute of Health and Care Excellence (NICE) to explore the use of representation of evidence statements in more efficient development of care guidelines, and this PhD project could build on or contribute to that.
Requirements: knowledge of (or willingness to learn about) first order and description logic, logic-based knowledge representation, ontologies; standard software development skills; clinical interest or knowledge is welcome but not required.